Research on CO₂ transfer in basalt and sandstone using 3D pore space model generation and mathematical morphology

Highlights

• 3D pore space models of basalt and sandstone are generated.

• Pore space characteristics are investigated using mathematical morphology.

• The potential CO₂ transfer path is numerically predicted.

• Simulation results are in accordance with results in previous studies.

Abstract

The performance of basalt and sandstone for CO₂ storage mainly depends on the CO₂ migration and solidification in the pore network of rock. This study provides points of view for the simulation of the three-dimensional pore space and the numerical prediction of gas transfer in the microstructure of the studied rocks. The biphasic change on Gaussian random field forms a randomly shaped pore network and the combination of biphasic fields creates a porous model that satisfies the real rock porosity and pore size distribution. Based on the generated model, several calculation scenarios using the mathematical morphology are proposed and numerically implemented to investigate the pore space characteristics, to extract the potential CO₂ transfer paths, to analyze the geometric features and thus to evaluate the CO₂ storage performance.

Introduction

Thermal power plant is the main power generation mode which produces about 85% of the world's electricity. It is impossible to totally replace the burning fossil fuels with clean energy that does not directly produce CO₂ in the short term (Wilberforce et al., 2019). As the most important environmental governance strategy, CO₂ capture and storage (CCS) technology was highlighted at the 21st Conference of the Parties (COP21) (Bui et al., 2018). Underground CO₂ storage is a permanent solution to alleviating emissions in the atmosphere and reducing the greenhouse effect (Oelkers et al., 2008; Wells et al., 2017; Schwartz, 2020). Approximately 8% of the world's continents and the majority of the seafloor are made of basalt, and these basalt layers have a huge potential for CO₂ storage (McGrail et al., 2006; Rosenbauer et al., 2012; Gislason et al., 2014; Snæbjörnsdóttir et al., 2020). When CO₂ is injected into active basalt, H₂CO₃ reacts with the divalent metal cations (Fe²⁺, Ca²⁺, and Mg²⁺) in the rock and solidifies, becoming part of basalt and thus is permanently stored (Matter et al., 2014; Snæbjörnsdóttir et al., 2014; Sigfusson et al., 2015; Snæbjörnsdóttir and Gislason, 2016):

$${\text{H}}_2\text{O}+{\text{CO}}_2={\text{H}}_2{\text{CO}}_3$$

$${\text{H}}_2{\text{CO}}_3+{(\text{Fe},\text{Ca},\text{Mg})}^{2+}=\left(\text{Fe},\text{Ca},\text{Mg}\right){\text{CO}}_3+2{\text{H}}^{+}$$

Besides the basalt layer, the deep saline aquifers of sandstone are also suitable CO₂ geological storage area (Xie and Economides, 2009; Varre et al., 2015). The Illinois State Geological Survey (ISGS) in USA conducted CO₂ injection experiments on Mt. Simon sandstone in Decatur area, and determined its feasibility for CO₂ solidification (Kohanpur et al., 2020). The Shenhua Carbon Capture and Storage (Shenhua CCS) project in China also performed CO₂ injection into Liujiagou sandstone layer, and achieved excellent storage effects (Ning et al., 2014).

As the basis for CO₂ migration, the three-dimensional pore network of rock affects the feasibility and efficiency of CO₂ solidification and storage. The connectivity of the pore space is the main factor. In order to conduct an in-depth study of this issue, scholars have used scanning techniques such as the micro-computed tomography (Micro-CT) to detect the real pore microstructure of rock (Xiong et al., 2018; Callow et al., 2018; Becker et al., 2019). Although direct imaging technology possesses the advantage of presenting the true morphology of material's microstructure, the selection of sample, observing preparation and scanning operation may bring inaccuracies. In particular, reliable and representative 3D photos with high resolution and sufficient scanning voxel size to comprise the whole range of pore size often lead to expensive cost and long duration experiment. Considering this, it is indispensable to reconstruct multi-scale digital models of rock based on the characteristic parameters of pore space (Joekar-Niasar and Hassanizadeh, 2012; Nomeli et al., 2014; Zhu et al., 2019). Traditional modeling methods of porous media usually define pores as simple elements with specific shapes, such as cylinders (Ranaivomanana et al., 2011) and spherical elements (Varloteaux et al., 2013). For instance, in recent years, Arshadi et al. (2020) used circular, triangular and rectangular cross-section forms to establish the pore network model of arkose. Kohanpur et al. (2020) used the maximal ball (MB) algorithm to equate pores in heterogeneous sandstone to spheres to study the two-phase flow of CO₂-brine. Song et al. (2019) used Avizo to equate pores to spheres to evaluate the permeability of rock. These methods gave us new ideas to investigate the microstructure and the related properties, but they are always based on the simplification of pore shape, thus are not able to present the actual randomness of pore form and spatial distribution.

Based on above considerations, in this study, the biphasic change on continuous Gaussian random field was numerically realized to generate the 3D porous model that satisfies the porosity and the pore size distribution of the studied rock. This method does not rely on an initial setting of pore shape and can present the random morphology of pore network. After finishing the model generation, numerical calculation scenarios based on the image analysis method of mathematical morphology were designed and applied to extract the potential gas transfer paths, to analyze the morphological characteristics (connectivity, tortuosity, pore throat, etc.) and thus to evaluate the CO₂ storage performance.